Hydroxynitrile lyase is an enzyme that catalyzes a reaction for producing cyanohydrins. Briefly, hydroxynitrile lyase is an enzyme catalyst used in the synthesis of cyanohydrins (α-hydroxynitriles) from carbonyl compounds in the presence of a cyanide donor. Cyanohydrins can be converted into various compounds and are useful as intermediates in organic synthesis. Further, optically active cyanohydrins may be used in the synthesis of α-hydroxyacids, α-hydroxyketones, β-aminoalcohols and the like, and are extremely useful in the synthesis of various important optical intermediates which are used for producing, for example, pharmaceuticals and chemicals. Therefore, development of a method for producing a large quantity of hydroxynitrile lyase is desired.
The hydroxynitrile lyase that catalyzes cyanohydrin synthesis is classified into the two groups of (S) selective group and (R) selective group. Of these groups, (S)-hydroxynitrile lyase catalyzes a reaction to produce (S)-cyanohydrin from either ketone or aldehyde and a cyanide compound under acidic conditions. As a representative reaction of this type, a reaction may be given in which (S)-mandelonitrile is produced from benzaldehyde and a cyanide compound, prussic acid. (S)-Hydroxynitrile lyase is also used as a biocatalyst capable of producing from inexpensive substrates optically active substances highly useful as intermediates for pharmaceuticals and chemicals. Thus, (S)-hydroxynitrile lyase is extremely useful in many fields. In order to utilize hydroxynitrile lyase in industrial production of optically active cyanohydrins, development of a method for producing a large quantity of a hydroxynitrile lyase which has high activity per cell or protein and high stereo-selectivity is desired.
It is known that hydroxynitrile lyase exists only in those plants that have cyanogenic glucosides. For example, as (R)-hydroxynitrile lyases, those derived from plants of the family Rosaceae, such as almond (Prunus amygdalus), are known. Further, as (S)-hydroxynitrile lyases, those derived from plants of the family Gramineae, such as sorghum (Sorghum bicolor); plants of the family Euphorbiaceae, such as cassava (Manihot esculenta) and Para rubber tree (Hevea brasiliensis); and plants of the family Olacaceae, such as tallow wood (Ximenia americana); are known. However, it has only been possible to extract extremely small quantities of hydroxynitrile lyases from these plants.
In order to obtain a large quantity of a hydroxynitrile lyase which is useful in pharmaceutical and chemical fields, attempts to obtain such a hydroxynitrile lyase with genetic engineering methods have been made(1-9). However, when a heterologous protein is expressed using a transformant, results obtained from researches on a corresponding homologous protein (such as yield, biochemical activities, etc.) are not always applicable. In other words, when a heterologous proteins is expressed using a transformant, it is not easy to predict the behavior and expression level of the transformant, the biochemical activities of the protein of interest, and so forth.
Further, depending on the type of the protein to be expressed, normal folding of the protein may not occur in the transformed host, resulting in the formation of an inclusion body. It is known that, in many cases, the protein within this inclusion body becomes an inactive protein without its inherent activity. In hydroxynitrile lyase, it is reported, for example, that 99% of hydroxynitrile lyase which has been expressed by culturing an Escherichia coli transformant at 37° C. is found in the insoluble fraction in the form of an inactive inclusion body(1). It is also reported that the enzyme activity of crude enzyme solution produced using E. coli is 0.545 units/mg protein(2-4); that the liquid activity of crude enzyme solution produced using E. coli (host: M15[pREP4]) is 0.5 U/ml(14); and that the specific activities of crude enzyme solutions are 0.20 U/mg protein (host: Top10′, 28° C.) and 0.61 U/mg protein (host: XL1-blue, 22° C.), respectively(15, 16). However, none of the above-mentioned enzyme activities is satisfactory.
As means to solve these problems, for example, an attempt has been made in which a hydroxynitrile lyase-expressing E. coli transformant is cultured at a low temperature to thereby inhibit the formation of inclusion bodies and improve the yield of a hydroxynitrile lyase having activity(8). However, this technology requires a long time for cultivation and needs to use large quantities of utility such as electricity and cooling water for maintaining the low temperature. Considering industrial production of hydroxynitrile lyase, these drawbacks will increase production cost greatly.
On the other hand, due to recent advancement in recombinant DNA techniques, it has become rather easy to prepare a mutant protein in which one or more amino acids constituting the original protein are deleted, added, inserted or substituted with other amino acids. In particular, when the protein of interest is an enzyme, it is known that mutants thereof acquire improvement in properties such as stability, resistance to organic solvents, thermal resistance, acid resistance, alkali resistance, substrate specificity or substrate affinity compared to the original enzyme, depending on the sites of the amino acid residues deleted, added, inserted or substituted and the types of the amino acids which replace those amino acids. These improvements in properties may bring about large reduction of production cost in industrial production utilizing enzyme reactions, through stabilization of enzymes as catalysts, simplification of reaction steps, improvement of reaction yield and so forth. Therefore, a large number of improved enzymes with various improved properties are now being developed.
In hydroxynitrile lyase, mutants in which one or more constituent amino acids are deleted, added, inserted or substituted have also been reported. For example, it is reported that a mutant hydroxynitrile lyase has an improved affinity to aromatic aldehydes, in particular, 3-phenoxybenzaldehyde(9, 10). However, a great increase in the production yield of hydroxynitrile lyase has not been achieved yet. It is also reported that a mutant in which tryptophan at position 128 is substituted with another amino acid and a mutant in which cysteine at position 81 is substituted with alanine were prepared and transformed into E. coli M15. When these E. coli M15 transformants were cultured in TB medium containing 100 μM IPTG under conditions cooled from 37° C. to 20° C., some of the mutant-expressing M15 transformants exhibited a hydroxynitrile lyase activity per cell higher than that exhibited by the wild-type hydroxynitrile lyase-expressing M15 transformants(6). However, according to this document, the hydroxynitrile lyase activity of the wild-type hydroxynitrile lyase-expressing M15 transformants is about ½ of that of the wild-type hydroxynitrile lyase-expressing JM109 transformants obtained under the same culture conditions. Thus, the effects of mutants are not necessarily demonstrated in hosts with high expression ability. It is also reported that substitution of glycine at position 113 with serine in a hydroxynitrile lyase derived from a subspecies of cassava (Manihot esculenta) grown in China increased the specific activity of the resultant mutant hydroxynitrile lyase(11). However, the amino acid at position 113 of the hydroxynitrile lyase derived from common cassava (Manihot esculenta) is serine. Thus, this report merely shows that this amino acid is important for hydroxynitrile lyase activity.
On the other hand, it is reported that the N-terminal methionine present at the time of translation undergoes processing in 40% of the proteins in E. coli cell extract(17). This processing is catalyzed by an enzyme called methionine aminopeptidase(18). It is reported that whether or not an endogenous protein in E. coli cells is ready to undergo processing by aminopeptidase is decided by the type of the amino acid at position 2 of the protein; and that the larger the side chains of the amino acid at position 2 is, it is more difficult for the protein to undergo the processing(12). There is also reported an N-end rule that “the stability of a protein in E. coli cells is decided by the type of the N-terminus amino acid of the protein”(13). According to this N-end rule, when the N-terminus of a protein is arginine, lysine, leucine, phenylalanine, tyrosine, tryptophan or the like, the stability of the protein in cells is low and the protein is readily degraded. Based on these findings, it may be possible to improve the stability of a protein of interest in a host transformant by selecting at position 2 of the protein an amino acid which has large side chains (i.e., hard to undergo the processing by methionine aminopeptidase) and which is not arginine, lysine, leucine, phenylalanine, tyrosine or tryptophan. However, the above-described results(12, 13) were obtained from analysis of endogenous proteins in hosts or some model proteins. Therefore, hydroxynitrile lyase which is a heterologous protein to a host will not necessarily follow the above-described rule because, as mentioned earlier, when a heterologous protein is expressed using a transformant, it is very difficult to predict the behavior and expression level of the transformant, the biochemical activities of the protein of interest, etc.
As described so far, attempts to obtain a hydroxynitrile lyase with remarkably improved properties can not be said successful. Creation of still more useful hydroxynitrile lyase mutants has been strongly desired.